Each month is a consistent win for AI. And July is no different.
While the duality of Artificial Intelligence is still very much in debate, the promise of business advancement, innovation and career opportunity are validated by amassing results.
What are some of the results?
July, alone, has us looking at progress in energy consumption, ways to reduce carbon emissions, and a robot dog that learned how to walk (in 60 minutes).
Faster Computation W/ Less Energy In Programmable Resistors
Training complex neural network models isn’t a cheap task.
Pushing the boundaries of AI and machine learning requires a lot of time and resources. And so a new area of research, known as analog deep learning, promises faster computation and reduced energy usage.
A team of MIT researchers are experimenting with protonic programmable resistors that use an inorganic material in the fabrication process to allow their devices to run 1 million times faster than previously. It’s also interesting to note that it is also roughly 1 million times faster than the synapses in the human brain.
Programmable resistors are also the building block of analog deep learning.
According to Science Daily, “These ultrafast, low-energy resistors could enable analog deep learning systems that can train new and more powerful neural networks rapidly, which could then be used for novel applications in areas like self-driving cars, fraud detection, and health care.”
As such, these resistors increase the speed at which neural networks are trained, while reducing the cost and energy to perform that training.
Lead author and MIT postdoc Murat Onen adds, “Once you have an analog processor, you will no longer be training networks everyone else is working on. You will be training networks with unprecedented complexities that no one else can afford to, and therefore vastly outperform them all. In other words, this is not a faster car, this is a spacecraft,”.
New Deep Learning Tool Will Help Cut Carbon Emissions
Speaking to your voice assistant in a slow, dragged out voice can confuse your device.
Often giving you results that you didn’t ask for. Or creating hindrances for the speech impaired.
University of Virginia cognitive scientist, Per Sederberg, suggests that current AI systems are hard of hearing. Which are problematic for both the user experience and for the environment.
UVA collaborative research has found a way to convert existing AI neural networks “into technology that can truly hear us, no matter at what pace we speak.”. The deep learning tool, SITHCon, can generalise input and understand words at different speeds to what the network originally has training on.
“This new ability won’t just change the end-user’s experience; it has the potential to alter how artificial neural networks “think” — allowing them to process information more efficiently. And that could change everything in an industry constantly looking to boost processing capability, minimise data storage and reduce AI’s massive carbon footprint.”
Robot Dog Learns To Walk (In 60mins)
A group of researchers at the Max Planck Institute of Intelligent Systems (MPI-IS) in Stuttgart were conducting research on how animals learn to walk.
So they built a robot dog to figure it out.
By learning to walk through sent and received sensor information, using reflex loops, and adapting its motor control, the robot dog was able to learn to walk in just one hour.
“Our robot is practically ‘born’ knowing nothing about its leg anatomy or how they work… The computer produces signals that control the legs’ motors, and the robot initially walks and stumbles. Data flows back from the sensors to the virtual spinal cord where sensor and CPG data are compared. If the sensor data does not match the expected data, the learning algorithm changes the walking behaviour until the robot walks well, and without stumbling. Changing the CPG output while keeping reflexes active and monitoring the robot stumbling is a core part of the learning process.”
Looking For More Content On AI?
Do yourself a favour and check out our blog!
It’s got tech news, tips and insights to take your digital enterprise to new heights.
If you enjoy getting a brief update on AI news, or love to read all about big data, digital transformation, machine learning and so much more – then be sure to check out the rest of our content…